汽车行驶工况特征参数优化研究

Research of Optimizing the Vehicle’s Driving Feature

提出一种汽车工况特征参数优化方法,通过试验车采集某城市的行驶工况数据,将预处理后的工况数据划分为工况段并提取出14种特征参数,综合考虑各特征参数之间的相关性、各特征参数与油耗的相关性以及各参数的变化速率和变化程度进行特征参数的优化,最终提取出平均速度和巡航时间比两个代表性特征参数。采用MATLAB神经网络模式识别工具箱,分别运用7类特征参数的组合方式对一段行驶工况进行识别。对比分析结果表明,优化后的两个代表性特征参数能够很好地对行驶工况进行识别,识别精度高、速度快,具有实用价值。

A kind of research for optimizing the driving features method is proposed in this article.Experimental data of a city driving cycle is obtained by driving test vehicle, dividing into " driving segments" after the pre-processing course, extracting 14 driving features from the driving segments. Optimize the number of driving features by considering the correlation between each driving features, the correlation between driving features and fuel consumption, the speed and scope of each driving feature change, and then obtain the typical driving feature as average speed and drive time percent. Use the neural network pattern recognition toolbox of MATLAB to recognize a test driving cycle by 7 driving features combination methods. Through the comparison and analysis, draw a conclusion that using the typical driving feature can recognize the driving cycle very well, having high precision of recognition and fast recognition speed, and the method have high practical value.